IJCAT International Journal of Computing and Technology

Volume 1, Issue 1, February 2014

www.IJCAT.org


115

Proxy Based Content Centric Networking with
Packet Loss Avoidance in Adhoc Network


1
Seenia Francis,
2
Anju John,
3
Divya Jose,
4
Pelja PauL.N

1, 2, 3, 4
Calicut University

Abstract - Rapid developments in the mobile technology have
transformed mobile phones into multimedia devices. Due to these
advancements, user created mobile content is on the increase,
both in terms of quality and quantity. In addition, content sharing
is getting popular in home networks as well as in social
community networks. To keep pace with such a movements the
new networking topology named as content centric networking
(CCN) optimized for content sharing has appeared. Due to
mobility of intermediate nodes, connection between two remote
nodes fails frequently bringing packet loss problem. Packet loss
cause communication delay, throughput degradation and
congestion and may even make the network unusable. This paper
proposes a new CCN scheme with packet loss avoidance built
into it.

Keywords Adhoc Network.

1. Introduction

A wireless LAN or WLAN is a wireless local area network
that uses radio waves as its carrier. The backbone network
usually uses cables. Some of the differences between
wired and wireless networks are these two types of
networks is one uses network cables and one uses radio
frequencies. A wired network allows for a faster and more
secure connection and can only be used for distance
shorter than 2,000 feet. A wireless network is a lot less
secure and transmission speed can suffer from outside
interference. Although wireless networking is a-lot more
mobile than wired networking the range of the network is
usually 150-300 indoors and up to 1000 feet outdoors.

Today, various contents are usually hosted by media
servers and web portals, and the only way to retrieve
contents is to establish an end-to-end connection with
them. While the communication pattern in the internet has
been evolving since its early days, from client-server
model to peer-to-peer networking, and to cloud
networking, one significant point has been shown, That is
The usage pattern of the internet has become largely
content oriented . That is, content consumers do not care
where and how to obtain a piece of content moreover, the
current internets connection exchange model leads to a lot
of signaling overhead, especially in the case of mobile
consumer devices, which introduces a whole range of
inefficient energy consumption. So the efficient
networking device has been considered by using a
radically different approach, namely content-based
networking where content queries and data are routed
based on content name.

CCN NODE MODEL

CCN is a new communication paradigm that has been
designed to substitute the current Internet. When compared
to the current TCP/IP communication model, CCN has the
following different characteristics:

i) Receiver-centric communication model: Receivers pull
information by sending an interest message. At most one
data message is delivered in response to an interest.

ii) Hierarchical content naming scheme: CCN does not
address specific hosts, but content object itself. Content is
given hierarchical names, which is similar to URLs.
Interest packets are forwarded by doing longest-prefix
Matching at forwarding decision phase.

iii) Cache and forward architecture: Every CCN devices
can cache data and use them to serve future requests

CCN communication is driven by the consumers of data.
There are two CCN packet types, Interest and Data (Figure
1). A consumer asks for content by broadcasting its
interest over all available connectivity. Any node hearing
the interest and having data that satisfies it can respond
with a Data packet. Data is transmitted only in response to
an Interest and consumes that Interest.1 Since both Interest
and Data identify the content being exchanged by name,
multiple nodes interested in the same content can share
transmissions over a broadcast medium using standard
multicast suppression techniques.

IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


116


Figure 1 CCN packet types

The FIB is used to forward Interest packets toward
potential source(s) of matching Data. It is almost identical
to an IP FIB except it allows for a list of outgoing faces
rather than a single one. This reflects the fact that CCN is
not restricted to forwarding on a spanning tree. It allows
multiple sources for data and can query them all in
parallel.

Figure:2 forwarding engine model

2. Related Work

2.1 Proxy-based Mobility Management Scheme in
Mobile Content Centric Networking (CCN)
Environments

The increase of user generated mobile content raises the
need of mobile content sharing. The proxy-based mobility
management in CCN environments that can provide low
control overhead and low packet loss. The goal of CCN is
to create a simple and flexible networking approach that
enables network to self-organize and relevant contents
where needed.

A CCN node (content requester, CR) asks for contents by
sending an Interest packet. Data are then routed the reverse
path back to the CR. The CR node has to ask for each
segment of the content in the same way. That is, one
content is composed of multiple segments. There are too
excessive control overheads during content sharing in
mobile CCN environments, especially when mobile device
is a content requester. That is, when the handoff event
happens, a number of redundant Interest packets have to
be sent again to retrieve the already-requested Data
packets. So, it results in long latency and too much control
overhead during node movement.

USER PROXY BASED MOBILE CCN SCHEME

This paper assumes that user proxies with CCN
functionality are configured as overlay architecture over IP
networks. So, if a user wants to get specific content data, it
just sends the content query (Interest) packet to its proxy
node. That is, user devices do not need to resolve and
make connections with other content holder devices by
themselves

Figure 3
.

Figure 4

The CCN proxy receiving the Interest packet may try to
discover the content itself using normal CCN Interest/Data
IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


117

exchanges. The proposed proxy-based mobile CCN has
the following operation procedures.

Step1. Handover (HO) detection: Mobile node detects
when to change network status by the use of physical link
information or router advertisements.
Step2. Handover indication: When detecting the handover
event is imminent, mobile node sends Hold request
message to its proxy node before the actual handover
event happens. After receiving Hold request message,
the proxy node stops delivering content Data packets to
MNs old location and only stores the content data in its
local repository for future retransmissions. After that, if a
Data packet is received, the proxy node checks whether a
specific PIT entry exists. Then, the received content Data
packet is not transmitted and just stored at the proxy
nodes repository. Therefore, the proposed scheme can
prevent unnecessary packet losses and network resource
consumptions toward the path to old location.
Step3. Handover complete: When acquiring new IP
address, the mobile node notifies the new IP address
information of its proxy node by using Handover
notification message piggybacking the content sequence
numbers that it finally received at the old location. The
CCN proxy node transmits the stored content data packets
toward the new location of mobile node. That is, the
proposed scheme does not need to transmit the repeated
Interest packets for the stored content data packets.
The advantages are provide lower communication
overhead ,provide low control overhead and low packet
loss,prevent unnecessary packet transmissions towards old
location during handoff.

The disadvantage is long latency and too much control
overhead during node movement.

2.2. Proxy-assisted Content Sharing Using Content
Centric Networking (CCN) for Resource-limited
Mobile Consumer Devices

This paper assumes that user proxies with CCN
functionality are configured in overlay architecture over IP
networks. The CCN proxy can be either a mobile device or
a PC that must be continuously active. The CCN proxy
behaves as the common point of routing path for all
content sharing. In addition, the proxy serves as a point of
indirection (providing relaying services) as well as
providing additional services such as message filtering,
secure association, and so forth. The basic idea is that user
devices ask the proxy to download the requested content
on their behalf (as shown in Fig. 5). The CCN proxy
participates in the conventional CCN overlay, and takes
care of all downloads of the devices. So, if a user device
wants to get specific content data, it just sends to its proxy
node the content query packet, which is similar with
interest packet in original CCN architecture. That is, user
devices do not need to resolve and make connections with
other content holder devices by themselves.

A. CCN overlay configuration

First of all, a mobile CCN device does a secure association
with a CCN proxy node for the content prefix
announcement and content sharing. This paper assumes
that the proxy nodes configure overlay network in advance
and therefore recognizes the identity information of others.
Each individual device including proxy nodes carries a
unique cryptographic identity in the form of a public key
pair. There is a tight coupling between the key pair used
by the device and its identity so that all devices can be
easily identified. Through the identity information, proxy
nodes can establish face configuration to construct routing
tables. After the secure association, the mobile CCN
device and the proxy node exchange the identity
information as well as IP address with each other. Storing
such information makes it possible to deliver content data
without exchanging additional interest packets when either
IP address of mobile consumer device or the serving proxy
node is changed. That is, once there is a secure association
between a mobile consumer device and one of proxy
nodes, there is no further processing at other proxy nodes,
except for network association (i.e., to create the face
configuration between them).


B. Proxy-assisted CCN content sharing

The mobile device that wants to receive a specific content
data sends a content request message containing the
requested content name to its proxy node
IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


118


Figure 6

The proxy node receiving the content request message
initiates a normal CCN content sharing procedure. That is,
the proxy node assumes the content request as one request
from its own application layer. After receiving the
metadata information of the requested content from
content holders, the proxy node delivers the metadata
information for the requested content data to the mobile
device. It is assumed that the metadata information can be
acquired when CCN content is generated and it is
piggybacked in the first segment data of the content. So,
the mobile device can configure the fake PIT entries for
the requesting content data without issuing further interest
packets in segment unit of the content data.

The advantage is prevent unnecessary packet
transmissions at the previous location that the mobile
device already moved out.It can save energy consumption
by reducing repeated transmissions of interest packets for
data packets not received during network change.

The disadvantage are proxy cost is high,when one proxy
failed it is difficult to receive the packet.

2.3 Mobility Management for Mobile Consumer
Devices in Content Centric Networking (CCN)

This paper proposes a partial extension based routing
update method for mobile content sources in CCN to
reduce network convergence time and the number of
routing table entry. The proposed scheme starts when a
mobile content source (MCS) sends a Prefix registration
(P Reg) message to its content router (CR) to inform the
movement event. At that time, a MCS locally announces
its name prefix to advertise its presence. CRs forward the
P Reg message towards the original domain. Through the
exchange of PReg messages, the extended path can be
configured from the original domain to the operation of a
partial path extension (PPx) scheme is largely made of 3
steps. The detailed procedure is as follows.

Step1. Movement indication: When detecting whether to
change network status, a MCS sends a P Reg message to
announce its presence. Here new domain of MCS to
provide seamless reach ability.
Step2. Path extension: The CR receiving a PReg message
compares its name prefix domain with that of the PReg
message.
Step3. Path update & revocation: In case that the MCS
moves away into another prefix domain networks, the
previously-established partial route has to be updated.

That is due to the fact that the route reach ability is
achieved in a short amount of time through a partial route
update that the extended route information is managed
only along CRs that constitute the path between the
original location and the new one. It can save resource
consumption of CCN networking by limiting the range of
routing update to the path between the original location
and the current location. The proposed scheme has lower
amount of overhead comparing with the basic CCN
scheme due to its partial route update nature. It keeps this
latency more or less constant until it reaches stabilization.

Contrary to client mobility, content source mobility creates
a complicated situation: a mobile source requests all
relevant content routers to update their routing tables for
the successful reception of future/on-going content
requests. However, it takes much time to update the
routing tables of all content routers. So, interest packets
may not reach, if the route to the relevant content source
changes; hence the interest packets are unnecessarily sent
again because content requesters do not know whether the
occurrence of interest retry event is brought about by
network problem or node movement. Furthermore, if
there are too many mobile content sources, it leads to the
pollution of the routing tables for their prefixes, which
counters the advantages of prefix aggregation.
Consequently, there may be a number of false routing
entries in CCN networks and then it leads to unnecessary
resource consumption and long communication setup
latency.

The advantages are the route reach ability is achieved in a
short amount of time through a partial route update that the
extended route information is managed only along CRs
that constitute the path between the original location and
the new one.It can save resource consumption of CCN
networking by limiting the range of routing update to the
path between the original location and the current location.
It keeps this latency more or less constant until it reaches
stabilization.The disadvantages are content source
IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


119

mobility creates a complicated situation. There may be a
number of false routing entries in CCN networks.

2.5. CSMA/CN: Carrier Sense Multiple Access with
Collision Notification

Under CSMA/CN, the receiver uses PHY-layer
information to detect a collision and immediately notifies
the transmitter. The collision notification consists of a
unique signature, sent on the same channel as the data. The
transmitter employs a listener antenna and performs
signature correlation to discern this notification. Once
discerned, the transmitter immediately aborts the
transmission.

Figure 7

While receiving a packet, the receiver uses physical-layer
hints to detect a collision and immediately notifies the
transmitter. The transmitter utilizes two antennas: one for
normal transmission and another dedicated to listening for
the notification. Upon detecting the notification, the
transmitter aborts its transmission, freeing up the channel
for other transmitters in the vicinity.

The operation of CSMA/CN can be summarized as
follows .The transmitter has two interfaces tuned to the
same channel, one for transmission and another for
listening. The receiver has a single interface (Fig. 1). Once
the communication begins, the receiver exploits preamble
correlation to detect the presence of an interfering frame.
Realizing that the packet reception is likely to fail, the
receiver checks the confidence of incoming bits via
physical-layer hints from Soft PHY. When the receiver is
reasonably confident of an error, it initiates a collision
notification to the transmitter. The notification is a short
signature unique to the receiver, also known to the
transmitter. The transmitters listening antenna
continuously searches for this signature using
correlation. We show that even in the presence of a strong
signal from the transmit antenna, signature correlation at
the listening antenna can reliably discern the collision
notification. The transmitter aborts, releasing the channel
for other nearby transmitters.
In CSMA/CN, the transmitter T uses one interface for
transmitting and the other (listener) for listening. The
receiver R uses its single interface for multiplexing
between transmission and reception. Transmission is
initiated as in IEEE 802.11, except one difference: For
every packet, the PHY-layer preamble is concatenated
with an additional bit sequence, a signature, uniquely
computed from the intended receivers identifier. T
ensures the channel is idle and transmits this packet using
the transmit antenna. The listening antenna, by virtue of
being very close to the transmitting antenna, receives this
signal with high signal strength (self-signal). The packets
intended receiver R also receives the transmitted signal
and starts decoding the arriving bits. Simultaneously, R
initiates collision detection.

Collision happens when a nearby transmitter T1 interferes
with Rs reception, causing packet corruption. To detect
such collisions, receiver R searches for a PHY-layer
preamble in its incoming signal. Searching occurs through
correlation of the preamble with the signal arriving at Rs
antenna. This happens in parallel and does not affect the
normal packet decoding procedure. Once T1s preamble
impinges on Rs antenna, the correlation exhibits a spike,
raising an alert that the packet may be in trouble. Arrival
of a new preamble may not necessarily cause a collision;
reception of the packet may be successful sometimes even
in presence of the interference. To verify the impact of
interference, R consults Soft PHY to obtain confidence
values of the bits arriving from T. The confidence value is
an indicator of how likely a bit is in error. Based on a
window of confidence observations, R infers whether the
packet is expected to get corrupted. If so, R halts reception
and prepares to send a collision notification to transmitter
T. The receiver R searches for a preamble while receiving
its frame of interest, but searches for its own signature
while receiving an interfering frame.

Upon detecting a collision, R stops receiving and prepares
to transmit a collision notification (CN). The CN is
composed of only Rs own signature. This is the same bit
sequence that T included in its packet to R. The receiver
transmits the CN packet like a regular 802.11 ACKthere
is no carrier sensing, hence the CN is transmitted even
though the transmitter is still transmitting. The listening
antenna of the transmitter continuously correlates for the
receivers signature in the incoming signal. This
correlation is more challenging because the self-signal is
much stronger than the notification .We show that even
then the listener can discern the notification with
consistent accuracy. Upon detecting the collision
notification, the listener immediately alerts the
transmitting interface, which then suspends the
transmission. Once the packet is transmitted, the
CSMA/CN receiver responds with an ACK when it is
IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


120

received correctly. However, unlike 802.11 ACK frame,
CSMA/CN ACK is simply a signature. If no ACK
signature returns from the receiver, the transmitter times
out and retransmit the entire packet.

CSMA/CN is an attempt to approximate CSMA/CD in
wireless networks. We show that it is feasible to abort an
unsuccessful transmission with the aid of a collision
notification from the receiver. Techniques from signal
correlation and Soft PHY-based hints are used.

The advantages are wasted transmissions are fewer in
CSMA/CN, resulting in better overall throughput.
CSMA/CN can detect most of the collisions at all bit rates.

The disadvantages are CSMA/CN cannot be used in
conjunction with multiple-inputmultiple-
output(MIMO).Traffic Conjunction


2.6 An Efficient Automatic Repeat Request
Mechanism for Wireless Multihop Relay Networks

Here relay stations are used to forward packets in the
networks.. When there are lost packets, relay stations
(RSs) decide whether to retransmit these packets with
automatic repeat request (ARQ) strategies. An improper
ARQ strategy increases latency, blocked packets, and
workloads on the multihop relay network. Here a new
relay ARQ (RARQ) scheme, providing efficient
acknowledgement to reduce packet latency and the number
of blocked packets with small workloads.

Compared with conventional single-hop wireless
networks, a relay network may suffer more packet losses,
thus increasing the overhead for handling packet losses. A
conventional way to manage packet losses is to use the
automatic repeat request (ARQ) mechanism.


Figure 8
This figure e(a),(b),(c) shows an overview of RARQ
scheme

Three conventional relay ARQ (RARQ) schemes that may
be applied in the relay network.

(a) The first scheme is the end to -end (E2E) RARQ, in
which all RSs simply relay packets, and the error
control is delegated to packet senders and receivers.
(b) The second scheme is the hop-by-hop (HbH) RARQ,
in which RSs are responsible for detecting errors, sending
acknowledgements, and retransmitting packets.
(c) The third scheme is the two-link (TL) RARQ scheme,
which divides an E2E path into a multihop relay link and a
single-hop access link, and a specific RS has to recover
packet losses for both links.

The disadvanteges are relay networks suffer more packet
losses.Increase the overhead for handling packet losses.

3. Proposed Scheme

A. Interest Packet Loss Avoidance

During content advertisement, intermediate nodes may
receive multiple paths that lead to the source node. In this
scheme all nodes keep record of the multiple paths in their
FIB tables sorted in terms of hop count and use them to
find alternative means to transmit packets. Other metrics
instead of hop count may be used. The next two sections
describe the two packet loss avoidance options for Interest
packet recovery.

1. Alternative Path Selection: When an Interest packet
arrives on a node, it first checks its cache. If the content is
not available in its cache the node then searches in its FIB
table for the next hop, with least hop count to destination,
and forwards it if it is available. But if the next hop node is
marked as unreachable, the node searches its FIB table
again and uses the next best available route to the
destination.

2. Broadcast: If all alternative next hop nodes are
unavailable or there is no alternative path towards the
content server, then the node broadcasts the Interest packet
with fixed TTL value to prevent flooding. TTL value of
two was arbitrarily chosen for this papers implementation.
Because of frequent mobility of nodes and change of
topology in wireless networks, new nodes that lie in a path
towards the content server but with no path entry in their
FIB table may arrive in the vicinity of the current node.
This broadcast scheme uses the newly arriving nodes as a
bridge to reach nodes in the path to destination server with
an existing entry in their FIB table. The nodes that
received the broadcast packet first check sequence number
and TTL value of the packet. If the TTL value is not zero,
the same content request with similar sequence number
was not received and the requested content isnt available
IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


121

in its content store, then the receiving node calculates
defer time and starts overhearing. After the defer time
expired and confirming that no other node has forwarded
the same packet, it decrements its TTL value, checks its
routing table and forwards it to next hop if available or
broadcasts it otherwise. Sequence number on data packet
is used so that any duplicate Interest packets that arrive on
a single node are not unnecessarily forwarded more than
once.

B. Data packet loss avoidance

In typical CCN when a node receives data packet, it looks
up the next hop node in its pending Interest table (PIT)
and forwards it. But if the next hop node is unreachable
and the link is broken, the node drops the packet and the
client will have to wait for its timer to expire before
sending its Interest request again.

After detection of link failure that causes data loss, our
proposed scheme uses broadcast to reach nodes that may
have pending Interest with same content name. Newly
arriving nodes or other intermediate nodes will further
broadcast the data packet to their neighboring nodes,
increasing the probability of reaching nodes that lie in the
path of the original Interest request. Similar to Interest
packet, the data broadcast packet also has sequence
number to prevent duplicate broadcast and fixed TTL
value to minimize flooding in the network.

The proposed scheme ensures before transmission that the
link with its next hop node is functioning before using it
and finds an alternative means. The proposed scheme
addresses two mobility scenarios that cause packet loss:

The first scenario is when an intermediate node moves
from its previous location. This causes both Interest and
Data packet loss. The proposed scheme avoids packet loss
by first using alternative path and broadcast. In the case of
Data packet, the scheme proposes broadcast since the Data
can only follow the path used by its associated Interest
packet. Even if the data is not received by the next node in
the path of Interest request, it will be cached in the nodes
that received the data broadcast and will delivered faster
next time it is requested by the client node.

The second scenario is movement of client node.
Movement of the client node that requested the data
affects delivery of data. The client will receive the data
using broadcast if it has not moved more than two hops
away from its preceding node. Even if it is more than two
hops away, the next time it requests the data, it will
receive it from nearby node since the broadcasted data
packet will be cached by the nodes in the data broadcast.

C. Movement indication

A mobile device detects when to change network status by
using physical link information or subnet address. In the
proposed architecture, both subnet change and proxy
change are considered to be a handoff event. As the
proposed scheme Content data packet towards mobile
devices. Content requester is directly related with the new


face. Configuration of CCN architecture to the proxy node.
In the same manner, the proxy change event should be
notified to the previous proxy node to prevent unnecessary
packet loss and resource consumption. When detecting
that the movement event is imminent, the mobile device
IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


122

sends Hold request message to a current proxy node.
After receiving the Hold request message, the current
proxy node stops delivering content data packets toward
the mobile device and only stores the content data in its
local repository for subsequent retransmissions. For that,
the HO field of the relevant entry in PIT is set to 1. As
shown in Fig. 7, 8 the CCN proxy node receiving interest
packet that indicates specific.

Content data configures content based routing entry for
future content data delivery. That is, the routing table
architecture of the proposed scheme has only difference in
the HO field of PIT configuration.

D. Prefix registration
( PReg) message -client node sends a prefix Registration
message to its proxy to inform the movement event.

E. Acknowledgement
When client node received data packet correctly, client
node sends an ack to proxy node delete the data and delete
the PIT entry.

F. Collision Notification

While receiving the packet the receiver uses physical layer
information to detect a collision and immediately notifies
the transmitter by a negative acknowledgement. Where
getting NACK,the proxy node retransmit the packet to the
client node.

In this scheme, mobile nodes detect availability of
neighbouring nodes using periodic overhearing of their
activities. If a neighbouring node fails to forward periodic
advertisement or it is inactive for some fixed time, the link
between the two nodes is considered to be unavailable.
The packet loss avoidance scheme begins when a
connection between two nodes is detected to be broken
this way. In the typical CCN, a node receiving an Interest
packet searches its FIB table for next hop entry. The
content provider sends the requested data by following the
reverse path of the Interest packet used. If the packet is not
received successfully, it will be dropped and the
intermediate sender doesnt take any prior action to avoid
occurrence of packet loss. The original node that requested
the packet will have to wait for some timeout value to
retransmit its requested interest packet Flow chart of the
proposed schemes packet


processing, Interest packet processing starts by checking
whether the packet is broadcast or uncast. If it is broadcast,
it checks its TTL value and sequence number and passes it
to the typical CCN's function of searching entry in FIB
table, otherwise it will drop it. If a next hop entry is
available for the content and there is no cached content, it
sends the packet using the best available route available. If
there is no next hop node, it calculates differ time, listen
its neighbours until timer expires and broadcasts it only if
no neighbour already sent similar packet. The unicast
packet processing follows similar procedure, except that it
doesn't need to check TTL value. In data packet processing
in our scheme. The additional features in this scheme are
that before deciding to send the data to its next hop node,
the current node makes sure its neighbour node is
available. If it is available it proceeds similar to the typical
CCN, otherwise it broadcasts the packet setting TTL value
of two and sequence number. In our proposed scheme
include backtracking and acknowledgement can be
provided to know whether the data is received or not.

The following section explains the proposed scheme for
recovering Interest and Data packet loss because of link
failure.

4. Performance Evaluation
IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


123


A. Simulation Environment

We used content centric network as defined by using
proactive routing update to populate its FIB table entry
and implemented using OPNET simulator. Our scheme
was also implemented using OPNET and compared it with
the typical CCN. We compared typical CCN and our
proposed scheme using four parameters, namely delivery
ratio, round trip time, cache hit ratio and prefix
announcement interval time variation. We used 15 seconds
as an interval period for prefix announcement by the
content server for the first three parameters. Table1 shows
the other parameters used in the simulation environment.

B. Result Analysis

Fig 9 depicts successful delivery ratio of packets measured
by varying node mobility speed from 0 to 20 m/s. Delivery
ratio is a ratio of how much of the packets requested by a
client node are delivered successfully.


As shown in Fig 9, the proposed scheme outperforms the
typical CCN. Initially at stable state, the typical CCN
performs slightly better than our scheme The reason for
that slight under performance is the usage of overhearing
causes false positive assumption of link breakage even if
the nodes are stable and the possibility of link failure is
low. That causes a node to use less optimal alternative
path or broadcast even if the best path is available. But
when we increase node mobility speed, the advantage our
scheme has over typical CCN becomes obvious and grows
to as much as 13% difference at 20m/s speed..

Fig . 9. Packet Delivery Ratio varying Node Speed

The second parameter used to examine the proposed
scheme is round trip time (RTT). The RTT is calculated
only for the successfully delivered user requested data. It
is an average time of all successfully served requests. Data
requests that were not delivered are not included in the
average RTT value. As shown in fig 10, the proposed
scheme consistently has higher RTT than typical CCN.
The proposed scheme has higher RTT value since, at the
event of link failure, the Data or Interest packets are sent
through alternate paths broadcast that generally take more
time than the original path. In the case of typical CCN, the
lost Interest or Data packets are not included in the
average RTT calculation since they are not successfully
delivered to the user. The more speed the nodes have the
better our proposed scheme performs, since there is more
link failure in higher speed.

Fig 10. Round Trip Time varying Node Speed
IJCAT International Journal of Computing and Technology
Volume 1, Issue 1, February 2014
www.IJCAT.org


124


Fig 11. Average Cache Hit Ratio varying Node Speed

Figures 11 show the performance metric in terms of
average cache hit. Consumers are randomly selected
among the mobile nodes in the simulated grid and the
speed of the nodes was kept between 0 to 20 m/sec
increasing in linear fashion. Advantage of the proposed
scheme becomes less visible due to effect of overhearing
and the subsequent use of alternate paths between the
nodes when the network is stable. Caching performance
difference between our proposed scheme and typical CCN
becomes more visible with increase of speed of nodes,
going as much as 20% improvement. With increasing
speed of the cache hit performance, our approach starts to
increase whereas in typical CCN the performance
degrades. When the nodes in a network increase their
speed, there will be more link failure and hence more
packet loss in the network.

Fig 12. Packet delivery ratio with packet announcement period variation

Figure 12 shows packet delivery ratio by varying content
providers prefix announcement rate. We started with a
constant node mobility speed of 10 m/sec, a prefix
announcement rate of 1 advertisement per 1 second and
increased it linearly up to 1 advertisement per 30 seconds.
The result in the graph shows that under heavy prefix
advertisement scenario of 1 advertisement per 1 second the
typical CCN performs better by sacrificing the network
performance due to heavy traffic being generated by the
content announcements. As we lower down the
advertisement interval gradually, we can clearly see that
our approach starts to perform better due to the fact that in
a high mobility environment low content announcements
result in stale FIB entries that dont reflect the recent
network setup. Hence typical CCN fails to forward packets
successfully since it is dependent on this stale FIB table.
But our proposed scheme recovers the failed links and
makes use of the newly arriving nodes to forward packets
to destination.

5. Conclusion

In this paper, we developed a new content-centric link
recovery scheme that improves content retrieval, caching,
and delivery in MANETs. The typical CCN
implementation waits for the client node that sent the
interest packet to detect packet loss, which increases
transmission delay and degrades throughput on the
network. Our proposal is designed to cope with the
wireless link impairments and highly dynamic topologies
and to limit the signaling overhead. Our proposed scheme
recovers lost packets by detecting the disconnected node
and trying alternative means to successfully transmit data
or interest packets. The link recovery is transparent to the
consumer and is performed without using any additional
control messages, which add overhead on the network.

Reference

[1] Van Jacobson et al. Network Named Content, in ACM,
2009.
[2] Soon Y. et al. CCN in Tactical and Emergency
MANETs, in Wireless Days Conference, Venice, Italy,
2010
[3] Michael et al. Listen First Broadcast Later: Topology-
Agnostic Forwarding under High Dynamics Technical
Report 100021, UCLA Computer Science Department,
2010.
[4] Marica et al. CHANET: A Content-Centric Architecture
for IEEE 802.11 MANETs, Network of the Future, 2011
[5] David B. et al. DSR: The Dynamic Source Routing
Protocol for Multi- Hop Wireless Ad Hoc Networks,
Carnegie Mellon University
[6] Perkins CE. et al. Ad-hoc On-Demand Distance Vector
Routing, AMCSA 99
[7] Marina et al. On-demand Multipath Distance Vector
Routing for Ad- hoc Networks, Proc. of 9th IEEE Int.
Conf. On Network Prot., 2001
[8] Charles E, Pravin Bhagwat Highly Dynamic
Destination-Sequenced Distance-Vector Routing (DSDV)
for Mobile Computers, SIGCOMM 94 London England
UK
[9] Jerry Wang et al. OLSR-R3: Optimized Link State
Routing with Reactive Route Recovery, in APCC,
Shanghai, China, October 2009